LUCIA AND ATREVIDA GLACIEBS 77 Terrace Point. In the latter place the glacier margin had lost its precipitous, broken ice cliff which was replaced by a moderately-sloping embankment of moraine-covered ice, steeper than in 1905, but ascended easily at practically all points. The margin stood almost exactly where it did in 1906 having melted back only a little, leaving a narrow marginal valley with no vegetation either in its bottom or for a short distance up the side of the terrace. Along the eastern margin, near Esker Stream, the jagged cliff was also gone and one could again ascend the glacier (PI. XXIII,) but up a much steeper slope, with thinner moraine cover than in 1905. The broken ice cliff above Esker Stream (PL XXI, B) was healed and the stream once more emerged from a well-defined ice cave (PI. XXII, A) though a much smaller one than that which existed previous to the advance. The stream volume was greater than in 1905 but less than in 1906. It was clear to us, though the evidence as to the exact amount is not conclusive, that this part of the glacier continued to advance for a short distance after it was last seen in 1906. This is supported by photographs in the two years showing the trees at the ice margin just west of the stream, where the ice had advanced many yards into the forest. Since the advance ceased there had been a slight recession, leaving a narrow barren zone like that at Terrace Point. At the base of Amphitheatre Knob there was a confused mass of broken and battered alder which was bombarded by falling bowlders and ice blocks during the advance; and in the debris that accumulated at the base of the ice cliff during this period, were incorporated the remnants of the alder thicket that grew on the glacier margin and in the marginal valley into which the ice cliff had been pushed. On the west side of Esker Stream, overturned and inclined spruce trees were seen in the morainic debris at the base of the glacier margin as they were in 1906; but instead of a jagged ice cliff rising above them (PL XXII, B), and constantly discharging bowlders, ice blocks, and torrents of muddy water and mud flows into the forest, there was a moraine-covered embankment down which small rock slides fell occasionally (PL XXIV, A). The destruction of this forest, so actively in progress during the summer of 1906, had ceased; but the effect of the advance on the fringing forest and on the alder thicket growing on the ice was strikingly apparent. There were hundreds of dead and dying spruce trees fringing the ice and partly buried in the morainic debris, and there were thousands of dead alder bushes, often in windrows, on the ice surface. In the piedmont ice bulb that spreads out beyond the mountain base was seen evidence of the same changes as have already been described for other parts of the area. It is clearly evident that the advance had about ceased when observed in 1906 and that the outer border of the glacier was unaffected by the advance. It, therefore, still retained its forest of alder, cottonwood, spruce, and hemlock. In the broken area the crescentic rings of breakage were still plainly visible but ice no longer appeared in them and no crevasses were seen. Ablation had so healed the scars that, viewed from a distance, the entire surface was a waste of ablation moraine; but there was a system of concentric crescents of ridges and valleys developed by ablation in the formerly broken area. The extent to which ablation had proceeded here since 1906 was vividly brought out by the destruction of the alder thickets. In 1906 the alder still grew on the inter-crevasse areas, and the crevasses formed crescentic white gashes in a field of green. Now the alder was all dead within the broken area and it lay in brown windrows piled up on the moraine-covered ice.